Examples

Runnable end-to-end code against the Revised API.

Complete, runnable programs rather than snippets. The guides show one call at a time; these show a whole job.

A Jupyter notebook that pages the directory with a cursor, flattens the banded metrics, ranks what comes back on transparent weights, reveals a name and exports an open-tier shortlist to CSV. Runs with or without an API key — it bundles 30 real open-tier listings so every cell executes offline.

Shorter snippets, in context

The guides carry working code for individual jobs:

The paging loop — cursor, page cap, rate-limit pause.
The band parser — "10-50", "<100", "250+", "1K-5K".
A weighted ranking that drops missing components rather than zeroing them.
The incremental sync loop — watermark, tombstones, idempotent upsert.
The republishing guard — in Python, pandas and jq.

Every example uses real endpoint paths and real field names. If one stops working, the API reference is generated from the live spec and wins.

Edit this page